This is an extremely important book. It is a study of 36 democracies covering the time span 1945-1996; shorter for some countries. The conclusion is that consensus democracies are better than majoritarian ones in a host of measures of the quality of the society, and are just as good by any economic measure.
The book is well written but I found reading it to be a challenge the first time through since I am not a political scientist; it was much easier the second time. The following is a summary. Quotations in the first three paragraphs below are from the Introduction.
"Defining democracy as 'government by and for the people' raises a fundamental question: who will do the governing and to whose interests should the government be responsive when the people are in disagreement and have divergent preferences? One answer to this dilemma is: the majority of the people. This is the essence of the majoritarian model of democracy."
"The alternative answer to the dilemma is: as many people as possible. This is the crux of the consensus model. It accepts majority rule only as a minimum requirement: instead of being satisfied with narrow decision-making majorities, it seeks to maximize the size of these majorities."
"The majoritarian model concentrates political power in the hands of a bare majority', though in practice it is merely a plurality, often less than a majority. "The consensus model tries to share, disperse and limit power in a variety of ways." "The majoritarian model of democracy is exclusive, competitive and adversarial, whereas the consensus model is characterized by inclusiveness, bargaining and compromise."
Characteristics of democracies are found to be in two groups of five, as shown in the table below. For each characteristic of the first group Lijphart finds a numerical value on what he calls the executives-parties dimension. Similarly each characteristic of the second group is found to have a value on the federal-unitary dimension. The grouping is entirely empirical. There is a high degree of correlation of the characteristics within a group; most correlations are significant at the 1% level, all are at the 5% level. Lijphart finds no significant correlation of characteristics of one group with those of the other group.
This raises the question of how can one find numerical values for characteristics that are not intrinsically quantitative? Lijphart, as others before him, has done his best. For example, how does one count the effective number of political parties in the legislature? One obviously does not given equal weight to a party with one member and a party with 100. A formula is used which seems satisfactory, though it is somewhat arbitrary. The final overall results will not depend on the precise formulae. Much of the book is a detailed discussion of the various characteristics and how they are quantified.
In order to find where each country lies between majoritarian and consensus democracy, Lijphart combines the values for the five characteristics on the executives-parties dimension by standardizing each to an average of zero and a standard deviation of 1, and averaging the five resulting values. The same was done for the federal-unity dimension. The executives-parties values are then plotted against the federal-unity values; it is found that they are uncorrelated.
The Table belongs here but I cannot figure out how to include it. Sorry.
Now comes the most important and interesting part of the book. The "so what" question. Does it matter whether one has a majoritarian or a consensus democracy? Conventional wisdom is that one needs a single-party majority government to make quick decisive coherent decisions that lead to better economic performance. It turns out this is wrong. Lijphart discovered that by looking at a range of economic indicators. There is no significant difference between majoritarian and consensus democracies, except that consensus democracies do significantly better in controlling inflation.
Then, Lijphart examines several measures of the quality of democracy and democratic representation. Consensus democracies come out better for the percentage of women in parliament, the percentage of women in cabinet, family policy (parent leave on childbirth, daycare, retirement system flexibility), rich-poor ratio, distribution of economic power, voter turnout, the satisfaction of voters with their democracy, distance between government policy and the wishes of voters, and two overall measures of democratic quality. These correlations are all significant at the 10% level, some at 5% and some at 1%.
Finally, Lijphart finds that consensus democracies tend to be 'kinder and gentler' than majoritarian democracies. Welfare policy, the degree to which despite unemployment, disability, illness or old age, people can maintain a decent standard of living through government spending is much better for consensus democracies. So is environmental responsibility and energy efficiency (GDP divided by energy consumed). Consensus democracies put fewer people in prison and are less likely to use the death penalty. Consensus democracies give a larger proportion of their GDP to non-military foreign aid. Again, all correlations are significant at the 10% level, some at 5% and some at 1%.
The conclusion is that consensus democracies are better than or as good as majoritarian ones on any economic or human measure. Are the correlations the result of cause and effect, or is it that an underlying consensual and communitarian culture leads to the choice of a consensus democracy? Lijphart suggests both are at work.
Lijphart concludes by stating that the overall performance record of consensus democracies is clearly superior to that of majoritarian ones. He then points out that it is not difficult to write constitutions, basic laws and other features of governance to introduce a consensus democracy, but that a change from a majoritarian democracy to a consensus democracy will require overcoming serious obstacles. Members of legislatures are very reluctant to change a system which got them elected. This is a trap. The only country to get out of it since 1945 is New Zealand in which, after two elections in which the party with the most votes lost the election, a Royal Commission recommended a change. The parties actually didn't want change but made promises of change they did not intend to keep. They were forced to keep the promises. It took two referendums to make the change from a British Parliamentary majoritarian system to a mixed-member proportional system; the first election under the new system occurred in 1996.